tylerhou

13 ms·

91.
by tylerhou 2y ago
No, I am not thinking of weak; this is why “inline does not mean inline” is on the iceberg :). Although practically speaking inline is also used as a hint. But chiefly it refers to inline linkage. I’m actually also not familiar with weak, b
92.
by tylerhou 2y ago
It’s more precise to say that inline means “this definition might be duplicated.” Non-inline functions also are allowed to have their definitions provided inline with the declaration; the difference is that the linker will error if it encou
93.
by tylerhou 2y ago
Flossing also hasn't been shown to not prevent tooth decay. (I.e. there is not much data to support one way or the other.) Also, some people don't carry the bacteria that create acid in their mouths, so (by chance) they don't
94.
by tylerhou 2y ago
No, this is not true. By default, starting a private chat with another employee defaulted to “history off” (as far as I can remember; I left Google >1 year ago now).
95.
by tylerhou 2y ago
C++ used to compile to C.
96.
by tylerhou 2y ago
Sorry if I am missing something, but why not run max-flow? Max-flow is simple to implement; basically just DFS/BFS in a loop. Also, a more advanced variant with better runtime, scaling max-flow, is not too much harder. https:/&#x
97.
by tylerhou 3y ago
C++11 was not legacy in 2013. It barely is considered legacy today.
98.
by tylerhou 3y ago
Based on is a bit loose -- it is indeed more precise to say that Datalog programs are often evaluated by database systems by translating rules into the relational algebra and running those queries until fixpoint. I can add that as an aside.
99.
by tylerhou 3y ago
A (set) relation with N tuples is isomorphic to a N-dimensional boolean tensor. The mapping is to interpret each tuple as a coordinate in the tensor, and set the entry in that tensor to 1.
100.
by tylerhou 3y ago
I'm lucky to have been surrounded by researchers who have been teaching and thinking about these ideas over the last ~6 months, so this essay has been marinating for a long time... your article just provided the necessary activation en
101.
by tylerhou 3y ago
Yes, you're right that one cannot express Datalog semantics (and also transitive closure semantics) with just one "query", as queries cannot be recursive. If you view each rule as a query, however, looping over rules does c
102.
by tylerhou 3y ago
Small nitpick: I would say that an abstract graph datatype does exist: it's a relation. I think that idea was not really properly explored by the referenced article. And the relational algebra gives us a powerful way of manipulating
103.
by tylerhou 3y ago
I'm confused by your link. The part where the paper said that they don't support page rank was referencing prior work on Cog. The system that they are now presenting seems to support PageRank queries. > However, despite the hig
104.
The "missing" graph datatype already exists. It was invented in the '70s(tylerhou.com)
335 points by tylerhou 3y ago|86 comments
105.
by tylerhou 3y ago
Even if there are technical limitations that prevent Apple from adding more I/O easily, as a consumer it feels like artificial segmentation/limitation. Apart from the initial M1 MacBook Pro release, it feels like most products App
106.
by tylerhou 3y ago
It depends on how risk tolerant you are. If you are at a startup, the growth of your startup is often largely dependent on how quickly you can add features / onboard customers. In that context, writing tests not only slows you down fro
107.
by tylerhou 3y ago
And the seismic retrofits aren’t cheap!
108.
by tylerhou 3y ago
0/1 ILP is NP-hard and the trivial algorithm takes O(2^N), and it's also in NP.
109.
by tylerhou 3y ago
https://people.seas.harvard.edu/~cs224/fall14/lec.html In particular, seems like lectures 9-11 have LP content.
110.
by tylerhou 3y ago
It was a response to > It's surprising how many problems can be formulated as linear optimization. i.e., all problems in NP (which is most problems you're likely to encounter on a day-to-day basis) can be solved with ILP, and m
111.
by tylerhou 3y ago
ILP is NP-complete.
112.
by tylerhou 3y ago
For regular service yes, but I've heard that sometimes they pull out the old cars for some reason.
113.
by tylerhou 3y ago
The authors spun it out into a startup, Feldera. A paper describing their idea also won Best Paper at VLDB 2023. The idea is very far from dead.
114.
by tylerhou 3y ago
I’ve seen it at stores where each item is relatively more expensive (e.g. a Uniqlo).
115.
by tylerhou 3y ago
Yes, there is a natural way to solve these recursively. Here is an example: https://leetcode.com/problems/best-time-to-buy-and-sell-stoc... Not the best runtime (I could use a multidimensional array instead of an hash
116.
by tylerhou 3y ago
Total, and for L3/L4. But this was my impression a few years ago; Google has changed comp a fair bit since then (e.g. vesting schedule).
117.
by tylerhou 3y ago
Canadian compensation for Google engineers AFAIK is about 60%-70% what it is in the Bay Area.
118.
by tylerhou 3y ago
In the branchless version, the CPU has to wait for the comparison to resolve before it can start executing the add for the next loop iteration. However, if the branch is predictable, the CPU can assume the result of the conditional and does
119.
by tylerhou 3y ago
If the value is Copy, you will make a copy and won’t take ownership, right? (Iterate over a vector of ints). Granted Rust structs are not Copy by default. Same with any C++ class for which there is no copy constructor.
120.
by tylerhou 3y ago
It’s a natural consequence of value semantics and not supporting references / pointers with range syntax :). (You can still write the same bug in Rust or C++ accidentally.)

More ›